Leaf Range in Colder Weather

My Nissan Leaf Forum

Help Support My Nissan Leaf Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
garygid said:
The lowest, weakest, least-charged cell-pair is what really determines when the car will stop. The pack voltage (sum of all the cell-pairs) can look good, like the battery has a good SOC, and the Pack Voltage could be used to calculate an "apparent" SOC estimate. However, one low cell-pair COULD stop the car long before this "apparent" SOC goes low.
Interesting theory - if that's the case - Nissan should probably adjust the SOC and DTE indicator to reflect the condition of the least-charged cell.
 
Since balancing needs to take place near the end of charge when the cells voltage starts increasing faster I wonder if charging to only 80% prevents full cell balancing. I know with LiFePO4 cells you need to get higher than 90% charge to do active top balancing of cells. Don't know the charge curve of LiMn cells well enough to say if that also applies.
 
garygid said:
Cost is not the issue here, Nissan already measures and monitors all 96 of the cell-pair voltages.

Just a software change needed, I think.

I don't think the cost is monitoring battery levels. I think the tough part and the expensive part is accurately determining the point at which charging is completed.

They attempt to minimize the risk by slowing down the rate of charge when the voltage reaches a certain level but the goal of hitting 100% without going over is like filling a wine glass in the back of a moving greyhound.
It is easy to say ok when we reach this we are full. But "this" is not a static figure.
 
It is quite easy to do, just slow (or stop) charging for a minute or so to get a better measurement of all the cell-pair voltages.

Then, continue charging (or not) at a suitable rate, scheduling the next "look" while still monitoring all the voltages for any cell-pair nearing a too-high condition.

By lowering the charge rate (under computer control), one should be able to avoid overcharging the highest cell, and begin cell "balancing" in a timely manner.
 
JRP3 said:
Since balancing needs to take place near the end of charge when the cells voltage starts increasing faster I wonder if charging to only 80% prevents full cell balancing. I know with LiFePO4 cells you need to get higher than 90% charge to do active top balancing of cells. Don't know the charge curve of LiMn cells well enough to say if that also applies.
One does not need to get higher than 90% to balance LiFePO4 - the voltage starts to rise at about 75%.

cc_cv_percent_1.jpg


Lithium is charged with a two-stage process - constant current, then constant voltage. Charge current drops steadily during the constant voltage state as the cells saturate. End of charge happens when the current drops to the manufacturer's set point - usually 0.1-0.2CA. One does not need to stop charging to balance the pack or detect the proper end of charge point.

One can fully charge and balance a LiFePO4 pack at 3.4V per cell just as easily as 3.6 or 3.7V.

charge_voltages.jpg



[Please note here that even when the cell was 'only' charged to 3.4V, the full CC/CV charge took place - and the charge terminated at the low current point (min I) at the end of the constant voltage stage. Please also note that we didn't lose much actual capacity even though the top and bottom voltage set points were well inside the factory max of 2.5-4.3V. In other words - we're not giving up much capacity to get a significantly longer battery life!]

The Leaf's BMS monitors voltages, temperature, cell pair internal resistance, charge/discharge current... It has all the info it needs to balance on the fly, or balance during charging, or adjust charging/balancing to temperature. It's much more capable than the simple analog units that many use very successfully.
 
garygid said:
<snip>
Further, it was just reported in another thread (I cannot remember which one) information (apparently from the Service Manual) that indicated that the PACK voltage IS used to derive the "SOC", not the "lowest cell-pair" information.

Then, this "apparent" SOC could be still relatively high, but the car's "protect-the-lowest-cell" monitoring MIGHT shut down the driving much earlier than expected.

IF TRUE, this could explain the several cases of "unexpected exhaustion", and the lack of the "Low Battery" warnings (assuming that they are indeed, as reported, based on this "apparent" SOC).
(buzzzzz) - thanks for playing! ;)

A couple of issues with this. First - the charge/discharge curve for LiMn is sloped like LiCo and lead-acid, not flat like LiFePO4. This makes it very easy to measure SOC directly from the cell voltage. This gives the car the direct ability to 'know' SOC of the total pack and each cell. Second - after reviewing three papers on cell production, it appears that 10-15% variation between cells is the norm. We might expect some variation in the cells in our Leaf battery, but we should expect that some cell matching has taken place so our variation should be smaller.

The combination of fairly consistent cell groups, years of experience and thus fodder for a decent pack model, and the ability to directly measure SOC suggests to me that even if the computer hasn't 'bonded' with it's new battery buddy, the car should still have the ability to tell us how far we can go on a tank of electrons.

From Leaf Service Manual, EV Battery (EVB-12):
The Li-ion battery controller is the core of battery control. This Li-ion battery controller detects the voltage and current of the assembled battery, the temperature of each module, and the voltage of each cell to judge SOC (state of charge) and calculates possible input/output values, meter indication value, and chargeable value to send these data to VCM (vehicle control module). VCM controls the vehicle, according to the battery state.

The pack is likely going to continue to change and may gain capacity during the first 50-100 cycles. The pack computer measures and stores cell internal resistance (at least) as part of it's process. It may take a couple of cycles for the car's computers to 'dial in' their pack info.

My plans for my Leaf, as of tonight, is to drive it home from the dealership in three hops of about 100 miles each. This will give the car three full (or nearly full) cycles to get good capacity data in the initial database. That should give the car better data to use when calculating range. We'll see how that works in May(be) or June-ish.
 
AndyH said:
JRP3 said:
Since balancing needs to take place near the end of charge when the cells voltage starts increasing faster I wonder if charging to only 80% prevents full cell balancing. I know with LiFePO4 cells you need to get higher than 90% charge to do active top balancing of cells. Don't know the charge curve of LiMn cells well enough to say if that also applies.
One does not need to get higher than 90% to balance LiFePO4 - the voltage starts to rise at about 75%.

cc_cv_percent_1.jpg
I guess it depends on the LiFePO4 cell since the CALB's we use have a flatter curve than that. More like this:
li_clip_image002_0006.gif

The voltage change is pretty minimal until you get near 90% or so. Maybe LiMn is different as I mentioned.
 
JRP3 said:
I guess it depends on the LiFePO4 cell since the CALB's we use have a flatter curve than that. More like this:
li_clip_image002_0006.gif

The voltage change is pretty minimal until you get near 90% or so. Maybe LiMn is different as I mentioned.
I'm familiar with the CALB curve, as well as TS, HiPower, A123-Systems, PSI, and a couple of LiPo/LiCo varieties. The PSI curve was stretched vertically. Here's a more typical view:

tssehp.jpg


The point isn't the main part of the curve, it's the rise above nominal near the end of charge. As I showed a couple of posts above, we can balance a LiFePO4 cell at 3.4V per cell just as easily as 3.7V per cell. (Also note that the charge and discharge curves are different - the charge in area 9 shows a distinct voltage rise even with a 3.4V charge that doesn't appear on the discharge in area 11. Most charts we see are discharge and do not give us enough useful info for charging/balancing behavior.)

It's exactly the same for LiMn as we get the same voltage rise near the end of charge - even if we finish at a lower voltage. Here's a series of curves for E-Moli LiMn2O4:

molirate.jpg
 
Balancing a SE/CALB cell at 3.4 is not going to be as accurate as doing so at 3.5 because the voltage is not as directly representative of SOC at 3.4. Real world my cells will take a good amount of charge at 3.4, meaning that I can have two cells both showing 3.4V but at different SOC. At 3.5V the cells will not stay at that voltage while taking more charge for very long. The further you get into the curve the more accurately voltage represents SOC. So yes you can balance cells at lower voltages but it will not be as accurate as going to higher voltages in the steeper part of the curve.
 
Thanks for the confirmation. ;) The good news is that the Leaf doesn't use LiFePO4. It might be better to move the CALB stuff to another thread.

The fact remains that lithium - and especially LiMn and LiCo - can be capacity balanced throughout the entire charge/discharge curve because of the sloped curve. This means that the Leaf doesn't have to wait to balance until the end of charge and comments about waiting until 80+ percent are irrelevant.
 
AndyH said:
Thanks for the confirmation. ;) The good news is that the Leaf doesn't use LiFePO4. It might be better to move the CALB stuff to another thread.

The fact remains that lithium - and especially LiMn and LiCo - can be capacity balanced throughout the entire charge/discharge curve because of the sloped curve. This means that the Leaf doesn't have to wait to balance until the end of charge and comments about waiting until 80+ percent are irrelevant.
I'm not sure I confirmed what you are suggesting. The point is that with LiFePO4 you really aren't doing much balancing at 3.4V since SOC can still differ. You need to get further into the end of the charge curve. Balancing is getting cells to the same SOC. Since 3.4V is not accurately representative of SOC at that point you aren't really balancing. They may get closer but it may not be close enough.
What you suggest about the LiMn LEAF pack having a steeper curve may be true, but I have yet to see a charge curve for the LEAF pack.
I'm not the only one questioning your assertion: http://www.mynissanleaf.com/viewtopic.php?f=27&t=2746&p=69018#p69018
Finally, 80% charging is not really at 80% SOC since Nissan limits the usable size of the pack. Charging at "80%" may actually be charging at closer to 70% since the full pack capacity is not accessed.
 
One can absolutely balance at 3.4V because I've done it. Works just as well and just as consistently as balancing at 3.6V, which I've also done (hmmm...and am doing right now. My bike is getting topped off at 3.6V per cell, and my solar storage pack is being balanced at 3.7V.)

Again - the Leaf is not a LiFePO4 vehicle so NONE of this applies!

Voltage is a second order indicator of SOC. And because of the relatively flat charge/discharge curve of some members of the lithium family, it's more difficult to positively track SOC until reaching one of the ends. But that's one of the main reasons manufacturers are using LiMn (aside from the price...) -- because the sloped charge/discharge curve provides an easy way to track SOC throughout the cycle.

Did you order a Leaf, JRP?
 
No, I'm sticking with my converted Fiero for now. It does what I need and I spent too much time working on it to give it up. Also, unlike the LEAF, it's available in my area ;)
 
Back
Top